Employing Artificial Intelligence to answer questions with your data
There can be huge opportunities for utilising the recent improvements in artificial intelligence (AI) to help make better decisions with data.
Increasingly large volumes of data are being processed and generated daily from multiple areas within the average business, and decision makers can find it challenging to glean any meaningful insights from the data.
Using pre-built AI features of Business Intelligence (BI) tools can help users explore data, find and extract patterns and then understand what the results mean. For example, dashboards could be created to predict how likely sales opportunities could be won or generate the predicted completion time for particular projects.
There are several AI based features and techniques now available in most modern BI tools such as Machine Learning, Natural Language Processing and Key Driver Analysis.
Depending on where you look, the definition for machine learning and artificial intelligence can vary. The definition I’ve found most useful is:
“Machine Learning is the science of getting computers to learn and act like humans do, and improve their learning over time in autonomous fashion, by feeding them data and information in the form of observations and real-world interactions.”
When a machine learning technique is used within BI software, this is generally used to “train” a data model using a set of data that will then generate a set of outcomes based on a collection of algorithms. This can be used to automatically generate a collection of insights across the dataset or provide predictions on future outcomes based on existing data.
For example, the Quick Insights feature available in Power BI can be used to automatically generate potentially interesting visualisations by scanning a dataset and discovering any patterns in the data.
Examples of insights generated could be the highlighting of any trends in time series data such as increasing costs or seasonality for particular business units, the correlation between the number of employees and total overtime costs when plotted against projects or identifying outliers such as employees with noticeably increased absence costs.
Natural Language Processing
Natural Language Processing, usually shortened to NLP, is a technology used to teach computers how to understand how a human communicates. Most NLP techniques rely on machine learning functionality to derive any meaning from human languages.
From a Business Intelligence perspective, NLP can be used to ask questions of your data simply by asking a series of questions. This could be used as an alternative to the more traditional filters and other interactive features present on a dashboard.
Continuing with Power BI as an example, the ‘Q&A’ feature is designed to help a user navigate through a dashboard by typing (or speaking) a series of questions and then interacting with the presented data to get the information they require.
Key Driver Analysis
Key driver analysis is a technique used to help understand what can drive particular business outcomes and then determine which factors impact key performance indicators. Many organisations use KPIs to measure success and sometimes the KPIs can be easy to identify, but sometimes they can be hidden within complex datasets and difficult to understand. The Key Driver analysis features within BI software uses machine learning models to scan the data, identify and rank any outcomes that could impact a KPI and then generates the results in a digestible format.
You don’t need to wait! These futuristic-sounding tools are available now and they are affordable and accessible for all businesses, no matter how small. With the help of AAB’s Systems & Software team, you could benefit from these evolving analytic & processing capabilities; saving you manual work and effort as well as delivering fresh perspective and insight into the success factors that drive your organisation.
By Stephen Lawie, Systems & Software Senior Manager at Anderson Anderson & Brown LLP
For more information about Stephen and the Systems & Software team, click here.